Parasitic Infections and the Immune System

Brazil.

Preface

In the past two decades the vast majority of (if not all) books and reviews dealing with the immunology of parasitic diseases have focused on host defense mechanisms, with an emphasis on how parasitic infections are contained and how the causative agents are damaged or destroyed by the immune system. Despite the severe immunologic alterations caused by many pathogenic parasites, which often impinge on the degree of pathogenesis and host survival, it has been a long time since a book has addressed the impact of parasitic infections on the host’s immune system.

Recent advances in immunology and other branches of biological science have opened exciting new avenues for exploring the mechanisms underlying the immunological alterations associated with parasitic infections. Consequently, a substantial body of literature has accumulated, particularly in the last 10 to 15 years, which has become increasingly more difficult to survey from original reports. In assembling the reviews composing this book, the editor selected contributions pertaining to major human parasitic infections. The term major reflects the large numbers of people affected or at risk worldwide as well as the public health and socioeconomic importance of these diseases.

It is anticipated that this book will be found useful or of interest by (a) scientists studying parasitic infections from a nonimmunologic perspective; (b) those concerned with the immunology of only some of the infections discussed herein and who wish to expand their knowledge and horizon; (c) newcomers to the field who grasp quickly the current state of the art in areas in which they wish to embark; and (d) parasitology and immunology teachers updating lectures on these major topics. The reviews should also enable scientists working in other fields to assess the applicability of their methods and systems to advancing the study of parasitic infections or, conversely, whether methods and systems used by others may help answer their own questions.

Of the eight chapters composing this book, six are devoted to malaria, Chagas’ disease (or American trypanosomiasis), African trypanosomiasis (or sleeping sickness), leishmaniasis, onchocerciasis and lymphatic filariasis, and schistosomiasis. These are the parasitic infections on which the World Health Organization’s Program for Research and Training in Tropical Diseases has placed its heaviest emphasis. The chapters on amebiasis and toxoplasmosis merit inclusion for their magnitude and world distribution. Recruited to the task were scientists recognized for their personal contributions to our current understanding of the impacts that these eight major parasitic infections have on the immune system. The unique characteristics of each parasite, their life cycles, their localization in the host, and the pathologies they cause can help us comprehend why the immunological impacts of pathogenic parasites vary so widely. Not surprisingly, then, the underlying mechanisms—to the extent that we understand them today—are often strikingly different. Therefore, and inevitably, the approaches used to unravel these mechanisms have varied dramatically for the different parasitic diseases, as clearly reflected in the pages of this book.

Parasites have learned to subvert the immune system to derive advantage while the immune system struggles, sometimes successfully, to reach a compromise on behalf of the host to ensure mutual survival. In the meantime, as witnessed in these pages, students of this compromise continue to pursue the goal of eventually attaining as much knowledge of immunology as parasites have unwittingly amassed over very long periods of time.

Felipe Kierszenbaum

1

African Trypanosomiasis

Maarten Sileghem, J. Norman Flynn, Ayub Darji, Patrick De Baetselier and Jan Naessens

Publisher Summary

This chapter describes African trypanosomes as protozoan parasites that cause disease in humans and livestock. Both Trypanosoma brucei rhodesiense and T. brucei gambiense cause sleeping sickness in humans. Tsetse flies only occur in Africa and Saudi Arabia. However, some trypanosome species can be transmitted in the absence of the tsetse flies and can be found far outside the African tsetse belt. Trypanosoma evansi only exists as bloodstream forms and is transmitted through mechanical transfer by biting flies. Most trypanosomes do not manifest a strict host tropism and can infect a variety of livestock species. They also infect wild animals, which form a reservoir from which the tsetse flies continuously reinfect livestock. The infected animals develop fever, lose weight, and progressively become weak and unproductive. Both Trypanosoma brucei gambiense and T. b. rhodesiense cause sleeping sickness in humans whereas T. b. brucei is unable to infect humans and is lysed by human serum in vitro.

I

Introduction

African trypanosomes are protozoan parasites that cause disease in humans and livestock. Trypanosoma brucei rhodesiense and T. brucei gambiense both cause sleeping sickness in humans. Other species such as T. brucei brucei, T. congolense, T. vivax, and T. evansi are not infective to humans, but all cause disease in livestock. At present 50 million people are at risk of contracting human trypanosomiasis. In addition, trypanosomiasis in livestock causes severe economic problems. In Africa alone, the widespread distribution of the tsetse fly, which is the vector for most economically important trypanosome species, makes 10 million square kilometers of potential grazing land unsuitable for livestock breeding (Fig. 1). Furthermore, roughly one-third of the cattle heard in Africa is presently at risk from the disease. Annual losses in meat production alone are estimated at U.S. $5 billion, according to the annual report of the International Laboratory for Research on Animal Diseases (ILRAD, 1989). This economic deprivation is exacerbated by a loss of milk production and a loss of tractive power. In this context it is appropriate to mention that T. evansi infects camels in Africa and in the Middle East and domestic buffaloes in Asia.

Figure 1 Distribution of tsetse flies and cattle in Africa. About one-third of the continent is unsuitable for cattle breeding owing to the widespread distribution of the tsetse fly. The impact of trypanosomiasis is even greater than this figure suggests because the areas inhabited by tsetse flies are potentially the most agriculturally productive in Africa. (Map kindly provided by Dr. R. Kruska.)

Tsetse flies only occur in Africa and Saudi Arabia. However, some trypanosome species can be transmitted in the absence of tsetse flies and can be found far outside the African tsetse belt. Trypanosoma evansi only exists as bloodstream forms and is transmitted through mechanical transfer by biting flies. The parasites are found in Asia, South America, and tsetse-free regions of Northern Africa. Trypanosoma vivax is transmitted by tsetse flies in Africa but can be found in South and Central America, where it exists in the absence of this vector. Trypanosomiasis is considered to be the major disease constraint on livestock development in Africa (Morrison et al, 1981b), but its importance is clearly not restricted to the African continent.

Most trypanosomes do not manifest a strict host tropism and can infect a variety of livestock species. They also infect wild animals, which form a reservoir from which the tsetse flies continuously reinfect livestock. Infected animals develop fever, lose weight, and progressively become weak and unproductive. Left untreated, many animals die from anemia, heart failure, and opportunistic bacterial infections. In humans, a similar course of events takes place, with parasites spreading into the central nervous system to cause the syndrome of sleeping sickness.

Figure 2 summarizes the life cycles for T. congolense, T. vivax, and T. b brucei, which is representative for the trypanosomes from the brucei group. The parasites live free in the blood and lymphoid tissues of the vertebrate host and are transmitted by tsetse flies. In all three species the parasites exist in the trypomastigote form in the vertebrate host. However, whereas T. b. brucei bloodstream forms are very pleiomorphic, this is not so for T. vivax or T. congolense. Following ingestion by tsetse flies feeding on an infected host, the bloodstream trypomastigotes transform to epimastigotes and later to metacyclic trypomastigotes. This transformation occurs in various locations depending on the trypanosome species (Fig. 2). In the vertebrate host, the parasites are covered by a surface coat that disappears in the tsetse fly and reappears on the metacyclic forms (Vickerman, 1978). The forms expressing a coat are shown in heavy outlines in Fig. 2. When the tsetse fly bites the vertebrate host, the metacyclic trypomastigotes are injected in the skin along with tsetse saliva, and a chancre develops at the site of bite. Chancre development, however, is not so marked in infections with T. vivax. The metacyclic trypomastigotes develop further in the chancre and transform into the bloodstream trypomastigotes, entering the local lymph vessels and later the bloodstream. The bloodstream forms may enter the connective tissues of the animal, although this is not usually observed with T. congolense. During human sleeping sickness, the parasites ultimately spread into the central nervous system.

Figure 2 Life cycles of T. b. brucei, T. vivax, T. congolense. The forms with a surface coat are shown in heavy outlines. [Reproduced with permission from the ILRAD Annual Report (1989).]

The morphological pleiomorphism of the bloodstream trypomastigotes seen during infection with T. b. brucei is associated with changes in metabolism. In the long slender forms, the mitochondrium is reduced to a peripheral canal, the Krebs cycle is not functional, and cytochromes are absent (Opperdoes, 1987). The parasites depend totally on glycolysis for their energy supply, and the NADH produced is reoxidized via a glycerol-3-phosphate oxidase system that is cyanide insensitive. In the short stumpy forms, which are nondividing differentiation forms (Shapiro et al., 1984), the mitochondrium is enlarged and fully active. This transformation is often considered as a preadaptation to the insect environment, where glucose is obviously not as abundant as in the blood of the mammalian host. The insect forms have an active mitochondrium. Glucose metabolism in bloodstream trypanosomes differs from glycolysis in other eukaryotes, and many of the enzymes involved in glycolysis are organized in specialized organelles named glycosomes (Opperdoes, 1987).

Trypanosomes from the brucei group are morphologically indistinguishable, and the different subspecies have been classified on the basis of geographical localization and host tropism. Trypanosoma brucei gambiense and T. b. rhodesiense both cause sleeping sickness in humans, whereas T. b. brucei is unable to infect humans and is lysed by human serum in vitro. The trypanolytic component of human serum has been identified as high-density lipoprotein (Rifkin, 1978). Since T. b. rhodesiense can be passaged through livestock species without losing infectivity for humans, sensitivity to lysis by human serum in vitro has been widely used as a more appropriate method for classification. However, some T. b. rhodesiense clones are able to switch from a serum-sensitive to a serum-resistant form and vice versa (Van Meirvenne et al., 1976) owing to the on–off switching of a single gene (De Greef et al., 1989). As a consequence, human noninfective forms of T. b. rhodesiense have regularly been classified as T. b. brucei. On the basis of isoenzyme patterns, it was found that T. b. gambiense is sufficiently different from the two other groups to be considered as a subspecies (Tait et al., 1984). However, T. b. brucei and T. b. rhodesiense are very closely related, and it is not clear whether they should be considered subspecies or variants of one species.

II Antigenic Variation

It was noted as early as 1911 that sleeping sickness is characterized by a succession of peaks of parasitemia in the blood. It is now well known that this wavelike pattern is due to the continuous appearance of trypanosome variants expressing a different antigenic identity (Gray and Luckins, 1976). The parasites are eliminated by host antibodies, but a few that have changed their antigenic identity are able to escape elimination and start a second wave of parasitemia. Again, new antibodies will clear the second wave, and parasites that have switched to a different antigenic type will continue the infection.

The appearance of different variable antigenic types (VATs) is caused by the expression of a different surface coat. The surface coat is constructed by one major glycoprotein, the variable surface glycoprotein (VSG). As a consequence, the succession of different VATs is related to the consecutive expression of antigenically different VSGs (Vickerman, 1978; Cross, 1975; Barbet and McGuire, 1982). Different VSG genes show extensive heterogeneity at the N termini but contain sequence similarities near the C termini (Rice-Fight et al., 1981; Matthyssens et al., 1981). When expressed on the membrane of living parasites, different VSG molecules are not cross-reactive. However, when purified, a cross-reacting determinant is detected (Barbet and McGuire, 1978; Barbet et al., 1981) which is located within the carboxyl terminal carbohydrate structure of soluble VSG molecules (Cardosa de Almeida and Turner, 1983). The carboxyl terminal amino acid from the VSG protein moiety is covalently bound to ethanolamine, which in turn is coupled to an oligosaccharide containing galactose, mannose, and glucosamine. This structure is bound to dimyristyl-phosphatidylinositol which anchors the VSG in the membrane (Holder, 1983; Ferguson et al., 1985a,b). When trypanosomes are disrupted, the amphiphilic membrane-form VSG is transformed into the commonly isolated water-soluble form (Cardosa de Almeida and Turner, 1983). This transformation is attributed to the action of an endogenous phospholipase C which cleaves dimyristyl glycerol from the glycolipid, releasing soluble VSG (Hereld et al., 1986). Cleavage of the phosphatidylinositol anchor is associated with the appearance of the cross-reactive determinant that is a cryptic epitope formed in part by the terminal inositol phosphate (Shak et al., 1988).

The switch of VSG molecules is attributed to gene rearrangements and transcriptional regulation. These include duplication of a basic copy gene followed by transposition to an expression site (Pays et al., 1981; Borst and Cross, 1982; Young et al., 1983a); in situ activation of telomeric genes (Laurent et al., 1983; Majiwa et al., 1982; Young et al., 1983b; Myler et al., 1984); recombinational chromosome end exchanges (Bernards et al., 1983); and transcriptional control of the expression site (Van der Ploeg, 1987).

The signal that induces the VSG switching is not totally understood. Antigenic variants appear in trypanosome populations cultured in vitro, indicating that the presence of antibodies is not a requirement to induce variation (Doyle et al., 1980). However, the immune system of the host plays a crucial role in establishing the wavelike pattern of the parasitemia and in defining the order of variant expression. Antigenic variation is not a random process, and new variants appear in an imprecise order (Miller and Turner, 1980). Mathematical models have shown that this order cannot be simulated on the basis of random generation and selection by growth rate (Kosinsky, 1980). More recently, a mathematical simulation model was proposed in which clearing by host antibodies and the existence of trypanosome forms with a mixed coat were both taken into account (Agur et al., 1989). Trypanosomes usually express only one VSG gene, but simultaneous expression of two genes resulting in the appearance of a mixed coat has been reported (Baltz et al., 1986). These forms might represent intermediate forms in the process of coat switching. According to the simulation model, the variant sequence is primarily determined by the antibody-specific mortality coefficient of the forms expressing a mixed coat and secondarily by the growth rate of the variants expressing a normal coat (Agur et al., 1989).

The number of VATs that can be expressed by a trypanosome population is very high. Around 100 variants were found to be expressed by a single clone using serological characterization (Capbern et al., 1977), and more than 1000 VSG genes were detected using DNA hybridization studies (Van der Ploeg et al., 1982). Thus, by undergoing antigenic variation, trypanosomes can escape destruction by the immune system almost indefinitely.

When the trypanosomes are taken up by tsetse flies, they differentiate into forms that no longer express VSG. At the end of the life cycle in the insect vector, the parasites transform into metacyclic forms that are infective for the vertebrate host. This transformation is associated with the reappearance of the variant surface coat, indicating that trypanosomes are protected by the VSG prior to contact with the vertebrate host (Vickerman, 1978; Gardiner et al., 1986). The metacyclic parasites in the salivary glands are antigenically heterogeneous but express a limited number of VATs (Le Ray et al., 1978; Hajduk and Vickerman, 1981; Turner et al., 1988). However, the metacyclic repertoire is unstable: a study of isolates from a focus of sleeping sickness in East Africa has shown gradual changes in the VATs expressed by the metacyclic forms over a 20-year period (Barry et al., 1983).

African trypanosomes live free in the blood of the vertebrate host and have no intracellular live stages. Therefore, they are a direct target for antibody-mediated destruction. It is thus obvious that antibodies are of major importance for the control of parasitemia in the infected host. Elimination of the parasites by antibodies in vivo is considered to be mediated by opsonization on liver macrophages (Mϕ) rather than by complement-mediated lysis (MacAskill et al., 1980; Dempsey and Mansfield, 1983a; Urquhart and Holmes, 1987). Because the surface coat overlies the plasma membrane and antibody-mediated trypanosome destruction is strictly VAT specific, the control of trypanosomiasis is attributed to VSG-specific antibodies. Anti-VSG antibodies were long thought to be T-cell independent (Campbell et al., 1978), but it has been shown that the VSG-specific antibodies generated during experimental infection represent a mixture of T-dependent and T-independent antibodies directed at different VSG epitopes (Reinitz and Mansfield, 1990).

In summary, the immunological control of the trypanosome growth is based on antibody responses to the VSG molecule. Owing to the variation of these molecules, trypanosomes are able to evade destruction by the immune system of the host. As a consequence, despite their importance in the control of parasite growth, VSG-specific antibody responses fail to generate a long-lasting immunity. At present it is not clear to what extent other aspects of the immune system are involved in the control of parasite growth and in the protection of the host from the pathogenic effect of the trypanosomes. This may in part be due to the drastic perturbation of the immune responsiveness following infection with African trypanosomes. Macrophages (Mϕ) are activated and CD5+ B cells are unusually expanded, whereas specific antibody production is impaired and T-cell proliferation is annihilated. It might thus be necessary to understand the impact of the parasitic infection on the immune system to gain a further insight into the complex interaction between the parasites and the host.

III Immunosuppression during Trypanosomiasis

One of the striking features of African trypanosomiasis is undoubtedly the profound suppression of the host immune system. Increased susceptibility to opportunistic infections during sleeping sickness was reported as early as 1903. Lobar pneumonia was commonly observed among trypanosomiasis patients (Low and Castellani, 1903), and streptococci were frequently isolated from patients who had died from the disease (Castellani, 1903). In 1973, a generalized immunosuppression was observed in trypanosomiasis patients that was proposed to be at least in part the cause of the increased susceptibility (Greenwood et al., 1973). Similarly, in livestock, infections with T. vivax and T. congolense cause suppression of antibody responses to a variety of vaccines (Scott et al., 1977; Whitelaw et al., 1979; Rurangirwa et al., 1979, 1982; Ilemobade et al., 1982). Other Trypanosoma species have not been studied in so much detail. However, it was reported recently that water buffaloes infected with T. evansi in Thailand have an increased prevalence of brucellosis (Bajyana Songa et al., 1987).

In view of the importance of suppression with respect to opportunistic infections, many investigators have turned to experimental infections in rodents to study the mechanisms involved in this phenomenon under controlled laboratory conditions. Suppression of antibody responses in laboratory rodents was first reported in 1970 (Goodwin, 1970) and has since been studied in detail. Various B-cell functions have been shown to be affected during trypanosomiasis, such as production of the immunoglobulins IgG and IgE to Nippostrongylus braziliensis (Urquhart et al., 1973), production of agglutinating antibodies following immunization with erythrocytes (Goodwin et al., 1972), in vitro release of IgM and IgG as measured by the hemolytic plaque-forming assay (PFC) (Murray et al., 1974; Hudson et al., 1975), and mitogen-induced proliferation (Corsini et al., 1977; Jayawardena and Waksman, 1977). Besides B-cell functions, many aspects of T-cell activity are severely suppressed. Experimental allergic neuritis (Altt et al., 1971), delayed hypersensitivity (Mansfield and Wallace, 1974), allogeneic graft rejection (Pearson et al., 1978), mitogen-induced proliferation (Jayawardena and Waksman, 1977; Pearson et al., 1978), mixed lymphocyte reaction (Pearson et al., 1978; Askonas et al., 1979), cytotoxic T-lymphocyte reaction (Pearson et al., 1978), antigen-specific memory cell proliferation (Gasbarre et al., 1981; Charoenvit et al., 1981), and secretion of migration inhibitory factor (Mansfield, 1978) are all blocked during infection. On the other hand, contact sensitivities to oxazolone (Murray et al., 1974; Urquhart et al., 1973) and dinitrofluorobenzene (Askonas et al., 1979) function normally. Furthermore, high titers of interferon-γ (IFN-γ) are found in the sera of infected mice (Bancroft et al., 1983; DeGee et al., 1985). IFN-γ is a cytokine normally produced by T cells and natural killer cells, indicating that some level of cellular activation occurs in vivo.

B cells are a direct target of the suppression (Eardly and Jayawardena, 1977; Askonas et al., 1979; Clayton et al., 1979a; Mansfield et al., 1981) but can also be affected indirectly through a modulation of the T-cell compartment (Mansfield and Bagasra, 1978; Selkirk et al., 1982). Most studies on suppression associated with experimental trypanosomiasis rely on infections with T. b. rhodesiense, T. b. brucei, or T. congolense. Suppression in rodents is also found during infections with T. b. gambiense (Oka et al., 1984), T. evansi (Yamamoto et al., 1985), T. vivax (Barrance and Hudson, 1986), and T. equiperdum (Oyejide et al., 1985), but these parasites have been studied less extensively.

IV Hypotheses Proposed to Explain Suppression in Rodent Models

During the early days of research on immunosuppression, it was noted that although the induction of specific responses is blocked, the B-cell population does not appear to be depleted or paralyzed but rather seems to be activated by the infection. The concentration of IgM in the serum is greatly increased (Houba et al., 1969; Neal et al., 1969), and the B-cell population is grossly expanded in vivo (Murray et al., 1974). It was therefore postulated by Urquhart et al. (1973) that trypanosomes possess a mitogenlike molecule which causes nonspecific activation of the B cells, leading to an inhibition of new responses. The existence of nonspecific B-cell activation was supported by the finding that autoantibodies and heterospecific antibodies reacting with a variety of trypanosome-unrelated antigens appear spontaneously during infection (Hudson et al., 1975, 1976; Askonas et al., 1979; Kobayakawa et al., 1979).

It seems rather contradictory that trypanosomiasis suppresses specific antibody responses to antigens in immunized animals and at the same time elicits antibodies to similar antigens in nonimmunized ones. However, B-cell activity is actually increased during the first days of infection and then rapidly drops to background levels (Hudson et al., 1976). This dual behavior resembles the pattern of specific antibody secretion in immunized mice treated with ordinary B-cell mitogens such as lipopolysaccharide (Diamantstein et al., 1976). As a consequence, it was postulated that trypanosomes cause a polyclonal activation which results in a progressive depletion or exhaustion of antigen-reactive B lymphocytes (Hudson et al., 1976; Askonas et al., 1979). Many investigators have tried to purify a mitogenlike factor from trypanosomes. In vitro studies on the mitogenic capacity of trypanosome lysates have yielded conflicting data (Esuruoso, 1976; Mansfield et al., 1976; Assoku et al., 1979; Selkirk et al., 1982). However, production of both polyclonal activation and immunosuppression is induced by in vivo administration of trypanosome membrane fragments (Clayton et al., 1979b; Sacks et al., 1982). Other authors reported a similar finding using purified VSG (Diffley, 1983). Interestingly, VSG immunization mimics the production of heterospecific antibodies and decrease in antibody response to immunized antigens but does not affect cellular proliferation. Spleen cells from VSG-immunized mice proliferate normally in response to both T- and B-cell mitogens, demonstrating that polyclonal activation is not linked to suppression of lymphocyte proliferation (Diffley, 1985).

The hypothesis that trypanosomes induce polyclonal activation links a variety of immune alterations to one central mechanism. The appearance of antibodies to foreign antigens, the increased serum immunoglobulin concentration, the massive plasma cell responses in vivo, and the reduction of specific antibody responses have all been related to polyclonal activation. Early studies on humoral responses during trypanosomiasis have also claimed that only a fraction of the total serum pool of infected rhesus monkeys was parasite specific (Freeman et al., 1970). This finding was later interpreted as an argument in favor of polyclonal activation. The idea that the presence of a mitogen might be the sole reason for a variety of immune perturbations is indeed attractive, but it is important to keep in mind that the close association of polyclonal activation and immunosuppression does not necessarily imply a causal relationship.

Other mechanisms have been proposed as the cause of immunosuppression, such as the generation of suppressor cells. These hypotheses rely predominantly on in vitro coculture experiments in which cells from infected mice are mixed with normal cells. Memory B-cell restimulation, secondary T-cell proliferation, and mitogenic activation of both B and T cells are all blocked in such cocultures (Corsini et al., 1977; Jayawardena and Waksman, 1977; Eardly and Jayawardena, 1977; Pearson et al., 1979; Wellhausen and Mansfield, 1979). Simulation experiments using in vivo VSG treatment have shown that spleen cells from VSG-immunized mice, in contrast to spleen cells from infected mice, are not suppressive in coculture experiments. However, both the induction of heterospecific antibodies and the inhibition of antibody responses to immunized antigens is mimicked following VSG treatment (Diffley, 1983). Thus, at least two different pathways, a suppressor cell-dependent and a suppressor cell-independent one, contribute to the overall suppression. As a consequence, polyclonal activation and suppressor cell generation are not necessarily two conflicting hypotheses but might represent coexisting mechanisms.

Subsequent to the discovery of suppressor cells, many investigators have attempted to characterize these cells phenotypically. In T. b. rhodesiense infection, the suppressor cells were found to be (a) absent from the T-cell-enriched nylon wool nonadherent fraction, (b) refractory to complement-mediated lysis of the Thy-1+ T-cell fraction, (c) depleted by adherence on Sephadex resin, (d) sensitive to talc treatment (phagocytic), (e) enriched in plastic-adherent fractions, and (f) present in athymic mice. On this basis it was proposed that the suppressor cell is a Mϕ-like cell rather than a suppressor T (Ts) cell (Wellhausen and Mansfield, 1979, 1980; Mansfield et al., 1981). In other models, the results were not so clear-cut, and different scenarios have been proposed. In mice infected with T. b. brucei, suppressor cells were proposed to be (a) present in nylon wool-purified T-cell populations, (b) depleted by complement-mediated lysis of the Thy-1+ fraction, (c) absent in peritoneal Mϕ populations, and (d) absent in athymic mice. Such data implicate the presence of Ts cells. However, adherent cells are also involved since suppressor cells are removed by glass wool adherence and are present in both plastic-adherent and plastic-nonadherent fractions. It was postulated that Ts cells can act directly but can also mediate suppression indirectly by arming adherent Mϕ (Jayawardena and Waksman, 1977; Eardly and Jayawardena, 1977; Jayawardena et al.,